DESCRIPTION (Applicant's Abstract): This is a request for a three year program of research on modifying the content and organization of an individual's perceptions and appraisals of self and others in order to improve clinical outcome in major depression. The major aim of this program is to adapt for clinical use research methods developed by the investigators (Gara et al., 1993; Goldston, Gara and Woolfolk, 1992) to study social perception and cognition in depression. These methods entail free-response descriptions of self and significant others by subjects and use a companion computer clustering algorithm (HICLAS: hierarchial clustering analysis) in order to represent how the free-response social perception data are organized at the level of the individual. In adapting these research methods for clinical use, we plan to simplify data gathering methods and presentation of structural (HICLAS) results so they can be readily used an assimilated by clinicians and patients alike. The use of these simplified methods is expected to provide feedback regarding the patient's implicit patterns of perception and belief about identity, significant others and important relationships, including the therapeutic alliance. The methods are viewed not as a stand-alone treatment but rather as an adjunctive treatment that will augment currently accepted treatments for depression, such as cognitive-behavioral therapy, that also focus on such patterns. Of particular interest in this work are the depressed person's thoughts and feelings about "self-as-patient," because prior work by the investigators has suggested that the degree of self-stigma that patients experience around this identity is a solid predictor of recovery from depression. Other aims at the project include full-computerization of assessment procedures and representational model, the writing of a manual to help professionals interpret and use HICLAS result therapeutically, and the development of criteria for rating therapist competence in using the methods and model. The project will also provide pilot data, using an experimental design, on the efficacy of HICLAS-augmented cognitive therapy (CT) versus CT without HICLAS.